[R] "New" problem with polr (or optim, or ...)

Anon. bob.ohara at helsinki.fi
Mon Jul 22 18:03:55 CEST 2002

Hello from a presently sunny Helsinki!

I've bee trying to repeat an analysis I did about 18 months ago (the
reviewers of the paper want something adding to it).  I'm using R1.5.0,
but I couldn't see anything in the list of changes between this and
1.5.1 to suggest it would act any differently.

The data is observational, on the changes in the population status of
carabid beetles, and various attributes of the species which may or may
not explain what's happening.  The response is ordinal - there are
several categories, from (roughly) "Gone Extinct" to "Suddenly all over
the damn place", and because the beetles haven't read Cochran and Cox,
the design is messy, with all sorts of confounding.  Now, when I run the
analysis I get this:

> ModelF <- polr(Ext ~ Habitat+Edge+Grids+SE+NH+Size+Wings+Grids:SE + Grids:NH + 
Habitat:Edge , na.action=na.omit) 
Error in optim(start, fmin, gmin, method = "BFGS", hessian = Hess) : 
        non-finite value supplied by optim

Which is presumably because the data is unbalanced.  This would be
(sort-of) OK, except that this particular model is the optimal one I
found when I ran it last time, using a previous version of R (and no, I
can't remember what version, sorry!).  So, (a) why doesn't this work
now, and (b) what can I do about it?  I've tried using simpler models,
but only models with one or two main effects work, and this is no good
for what I want.  Any help would be appreciated!

Thanks in adavance.


Bob O'Hara

Rolf Nevanlinna Institute
P.O. Box 4 (Yliopistonkatu 5)
FIN-00014 University of Helsinki
Telephone: +358-9-191 23743 
Mobile: +358 50 599 0540
Fax:  +358-9-191 22 779
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list